Flow out of the system within 2D area

[Anonymous]

Hi,

I am modelling a dam breach with a network of interconnected canals downstream. It has been easier modelling the whole area with as 2D rather than connect 1D to 2D elements. However, at the end of one of the canals the water flows into a pipe that is quite long and end at a completely other location. I don’t want to and see no good way to model the underground pipe, so I created a small storage area and connect it to the 2D flow area. When I build the model, I could not connect the cells in the middle of the mesh with the storage area, but only the cells at the boundary of the 2D area. This forced me to create two 2D areas and connect them with 2D area connectors. The 2D areas have exactly the same outlines because I copied the coordinates. The connection is also made with the coordinates, so it should be a seemless transition. But it considerably slows down the model to a run time of about 3 hours. Is there a better way to model flow out of the system in the middle of a 2D area? Do I have other options than what I did?

Thank you for your help
Regards,
Stephanie Morach

[cameron]

This is an issue that HEC is aware of and source/sinks inside a 2D domain did not make it into the 5.0 release, but it is on their radar and will be included in a later version.

As for how you modeled it, did you have to change the time step which caused to to run slower or is it hitting max iterations? There are a few 1D/2D connection options you can adjust that may help.

[Anonymous]

Hi Cameron,

Good to know that it is on their radar, and it will make HEC-RAS even better.

I did not adjust the time step when inserting the connections. It runs fine until the flood wave arrives and from then on I have a 1D/2D flow error every time step, even as the flows go down to a very small value. It is slowed down because of the max iterations.
I am currently modeling the connection with the following parameters:
– the 2D area boundaries are exactly on the same place/line
– I use the normal 2D equation domain
– I copied the weir data from the terrain profile as given in the connection editor, and adjust the few nodes where HEC-RAS give me an error message that the elevation is lower than the connected cell
– In the HTab Parameter I have the number of points/curves set to the maximum value of the range and the head water maximum elevation is a reasonable value that is a bit higher than what happens, but not lower and not exceedingly higher
– I have tried computing it with both “compute flow each time step” as well as “compute as curves (faster)” and it hasn’t made a difference

I hope this information helps a bit. I appreciate any help what 1D/2D connection options I can change that may help.

Thanks for your help
Stephanie

[cameron]

If you go the options and tolerances options in the unsteady plan pull down menu, there is a 1D/2D options tab. Set the maximum iterations to something other than zero, you will have to test out what works best for your model. Setting it to 20 would make it the most stable probably, but will add to the run time.

[Author]

Posts